navi-fractal¶

Audit-grade fractal dimension estimation for graphs.

Zero runtime dependencies. Refuses to emit a dimension without positive evidence.

Measure in five lines¶

from navi_fractal import make_grid_graph, estimate_sandbox_dimension

grid = make_grid_graph(30, 30)
result = estimate_sandbox_dimension(grid, seed=42)
print(f"D = {result.dimension:.3f}")  # D = 1.620

The refusal is the feature¶

If your network doesn't have fractal structure, you don't get a number --- you get a machine-readable refusal with the exact reason why.

Complete graph K50    → Refused: trivial_graph
Barabási-Albert       → Refused: no_valid_radii
Erdős-Rényi random    → Refused: no_valid_radii
(1,2)-flower          → Refused: no_valid_radii
30×30 grid            → D = 1.620, R² = 0.9999
(2,2)-flower gen 8    → D = 1.810, R² = 0.9994

Quality gates, not warnings¶

Every result passes through a chain of statistical gates before emission:

Gate	What it checks
\( R^2 \) threshold	Power-law fit must actually fit
AICc model selection	Power-law must beat exponential decisively
Curvature guard	Reject windows where quadratic fits better
Slope stability	Reject windows with high local slope dispersion

If any gate fails: dimension=None, reason=<why>. Every result --- accepted or refused --- is a frozen dataclass with the full audit trail.

Calibrated against ground truth¶

Sandbox estimates are validated against (u,v)-flower networks with exact analytical dimensions, proved in Lean 4 by fd-formalization.

See the calibration table →

Citing this project¶

@software{spence2026navifractal,
  author = {Spence, Nelson},
  title = {navi-fractal: Audit-grade fractal dimension estimation for graphs},
  year = {2026},
  url = {https://github.com/Project-Navi/navi-fractal}
}

License¶

navi-fractal is released under the Apache 2.0 License.